At its annual Think conference, IBM announced a new Watsonx platform that provides tools for creating artificial intelligence models and access to pre-trained AI variations capable of generating computer code, text, and much more.
IBM says that the launch of the new platform is partly due to the problems that many firms face in the process of introducing artificial intelligence into the work environment. The company conducted a survey, according to which it became known that 30% of business leaders are in no hurry to integrate AI into their commercial structures due to trust issues and concerns about the transparency of the technology. Also, 42% of respondents refrain from active use of generative artificial intelligence, referring to the unresolved issue of confidentiality.
Rob Thomas, IBM’s commercial director, said that the probability that AI will replace managers is not the maximum, but it is certain that specialists of the appropriate profile using advanced technologies will replace those who do not.
Watsonx provides clients with access to a set of tools, infrastructure, and consulting resources necessary to create their artificial intelligence models or accurately tune existing AI configurations based on their data. The company characterizes the new platform as a corporate studio for artificial intelligence developers. Users will have the opportunity to test AI models and deploy them, tracking the subsequent functioning of the technology.
Amazon has a similar product, SageMaker Studio, and Google has Vertex AI. Azure also has an artificial intelligence platform of the same name. IBM claims that Watsonx differs from competitors’ products in that it provides several pre-trained AI models and a cost-effective infrastructure.
Dario Gil, senior vice president of IBM, says that the company offers horizontal opportunities at a time when there is still a need for a large organization and team to innovate in the field of artificial intelligence.
Watsonx offers seven pre-trained AI models, some of which are open-source. The company has also partnered with the artificial intelligence startup Hugging Face to gain access to thousands of models, datasets, and libraries.
The AI configuration that IBM highlights are fm. model. code generating code, fm. model.NLP, a collection of large language models, and fm. model. geospatial, a variation based on NASA climate data and remote sensing.
Fm.model.code provides an opportunity to formulate a task in natural language, which will then be automatically implemented in the coding workflow. Fm.model.NLP contains functions for generating text materials for specific thematic groups. Fm.model.geospatial can predict changes related to natural disasters, the transformation of biodiversity, and land use. Forecasting is also available for several other geophysical processes.
IBM claims that the models were trained based on many different data sets, including business information, tabular data, geospatial data, and IT event data.
Arvind Krishna, CEO of IBM, stated that the company allows enterprises to use their code to adapt the AI configuration to specific needs.
As we have reported earlier, IBM Suspends Hiring Due to AI.